Improving ML Attacks on LWE with Data Repetition and Stepwise Regression
Alberto Alfarano, Eshika Saxena, Emily Wenger, François Charton, Kristin Lauter · 2026
Read Improving ML Attacks on LWE with Data Repetition and Stepwise Regression by Alberto Alfarano, Eshika Saxena, Emily Wenger, François Charton, Kristin Lauter (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-03903.txt.
Description
Alberto Alfarano and colleagues strengthened machine-learning attacks on Learning with Errors, a hard lattice problem behind much post-quantum cryptography. Earlier attacks recovered only sparse secrets, with up to 3 active bits in the cruel region on BKZ-preprocessed samples. Larger training sets plus repeated examples let the team recover denser secrets, following a power law linking attempts, data size, and repeats. A new stepwise regression technique then recovered the cool bits of the secret.
Catalog record
| CREATOR | Alberto Alfarano, Eshika Saxena, Emily Wenger, François Charton, Kristin Lauter |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 51962 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.03903